This is an example of a very basic on-chain monitoring and alerting. The idea is to provide a basic structure, which can be simply expanded while keeping the complexity of every part at a mininum.
There are two different types of scripts: check scripts that get the needed data from an api endpoint, parse and check the retrieved values and trigger the notification script(s), if necessary. The notification scripts just contain the commands needed for sending the alert.
Of course you don't even need to use API endpoints at all and retrieve the data with your cosmos-sdk binary locally. By using an external API, you can even let your monitoring run on seperate hosts.
Just put your monitoring scripts into a crontab
0 */4 * * * /path/to/check-funds.sh
*/15 * * * * /path/to/check-oracle-misses.sh
You probably need to adjust the interval for every script individually.
You can add an alarming endpoint by simply adding a new script. If you want to have notifications to different Discord channels for example, you could have multiple scripts like send-discord-general.sh and send-discord-important-news.sh, each with a different url and use them seperately.
Or you modify the notification scripts and pass parameters instead of creating copies.